Measuring Racial Disparities in Police Use of Force: Methods Matter

Autor: Phillip Atiba Goff, Dean Obermark, Tracey Lloyd, Amelia M. Haviland, Amanda Geller, Jack Glaser
Rok vydání: 2020
Předmět:
Zdroj: Journal of Quantitative Criminology. 37:1083-1113
ISSN: 1573-7799
0748-4518
Popis: To understand the impact of measurement and analytic choices on assessments of police use of force (UOF) and racial disparities therein. We collected and standardized UOF data (N = 9982 incidents) from a diverse set of 11 police departments, and measured departments’ aggregate force severity in five ways. We assessed the sensitivity of racial disparities in UOF severity to a series of analytic choices, using a 5 × 2 × 2 × 2 design comparing force severity to population and arrest benchmarks, using two definitions of minority group (Black/Nonwhite), and two modes of comparison (ratios/differences). Significant racial disparities were observed under most analytic choices in most departments. However, lethal force was rare, and estimates of lethal force disparities were statistically uncertain, as were departments’ relative ranks as equitable or disparate. Ratios of minority to White force severity were less sensitive to measurement differences within measures including nonlethal force. The choice of a population or arrest benchmark had implications for which departments emerged as highly disparate, while focal minority group and mode of comparison had less systematic effects. Given increased scrutiny of police activity by advocates and policymakers, it is important to understand how measurement and other analytic choices affect our understanding of equity in police practices. Our findings demonstrate that analytical decisions interact in complex ways and that standardization is essential when comparing multiple departments. We recommend comprehensive data collection that includes nonlethal as well as lethal force, and make recommendations for measuring and contextualizing racial disparities in UOF and other police activity.
Databáze: OpenAIRE